OBJECTIVES: Baseline characteristics of patients with focal seizures recruited into adjunctive AED trials have evolved over time, becoming more refractory and severe. Concurrently, placebo responses have increased. This analysis used propensity-score matched patient-level data in an indirect comparison of brivaracetam and levetiracetam to account for potential sources of heterogeneity. METHODS: Patient-level data from randomized placebo-controlled trials of brivaracetam (recruited 2007-2014) and levetiracetam (recruited 1993-1998) were pooled. Consistent inclusion/exclusion criteria were applied. Potentially confounding baseline characteristics were matched using propensity score weighting and outcomes were defined consistently, thus reducing the potential for bias. Placebo response was used to benchmark matching success. RESULTS: After applying inclusion/exclusion criteria, 707 and 473 active drug and 399 and 253 placebo patients comprised the brivaracetam and levetiracetam groups respectively. Before weighting, there were significant differences between groups for demographics, seizure type and frequency, epilepsy history, number of concomitant AEDs, comorbidities and etiology. Median percent seizure frequency reduction in the brivaracetam-placebo arm was 21.7%, and 3.9% in the levetiracetam-placebo arm. After weighting, the patient groups were balanced on all observed baseline characteristics; however, median seizure frequency reduction was still 15.0% in the brivaracetam-placebo arm and 6.0% in the levetiracetam-placebo arm. CONCLUSIONS: The propensity-score matching did not successfully adjust for placebo response differences, though there was a minor improvement. Since patients were matched on observed potential confounders, this would indicate the presence of unobserved confounding factors associated with placebo response. As such, adjusted or unadjusted indirect comparisons between brivaracetam and levetiracetam using available trials remain problematic. Previous research has indeed found that variation in placebo response over time in AED trials is not fully explainable by observed baseline characteristics. Future trials, and subsequent indirect comparisons, should better account for the heterogeneous nature of epilepsy and the need for individualized care, to have greater value in healthcare decision making.